Latest / Catholic Spirit Radio / Being Catholic #374: AI: A New God or Just Hype?

Transcript

Being Catholic #374: AI: A New God or Just Hype? · Catholic Spirit Radio

0:00
This is Being Catholic with Bob Johnston on Catholic Spirit Radio.
0:06
Hi, this is Bob Johnston, and you're listening to Being Catholic,
0:10
right here on Catholic Spirit Radio, 89.5 FM and 92.5 FM in good old McLean
0:17
County in Bloomington Normal,
0:19
88.3 in Pontiac, 97.1 in Lincoln 89.3 in Morris Joliet 89.1 in DeKalb Sycamore
0:29
88.3 in the Rockford area Marengo, Harvard,
0:33
Beloit 89.3 in Whitewater, Wisconsin and 88.3 in Clinton and Champaign,
0:40
Illinois covering much of central and northern Illinois also southern Wisconsin
0:45
and still growing thanks to you we're going to have a great show for you today,
0:50
I'm here with my wife, Lynn, and we are going to talk about AI today.
0:56
There is a lot of hype about AI and a lot of fear about AI and a lot of ideas
1:02
actually that bear on Christianity that somehow or another AI will replace God
1:08
or cast us into a godless society and so forth. And so we're going to address some of those things.
1:15
But before we start, we'll talk a little bit about our new Pope,
1:19
and I'm going to turn this over to my wife and see what she has to say about that. Okay.
1:25
This has been such a tremendous week in a sense.
1:31
It's overwhelming. I can't hardly still believe that we have a Pope from America,
1:37
from North America, from Chicago, and the south side of Chicago.
1:44
Hey, you think he's a Sox fan, a white Sox fan, or a Cubs fan?
1:51
Well, he said to be a Sox fan, but then I heard just last night that there's
1:57
some doubt that he may be a Cubs fan. So I'm not sure.
2:00
I would guess he would be a Sox fan, wouldn't you? He's from the South Side. Yeah.
2:05
You would think, surely, if you're from the South Side, he'd be a Sox fan.
2:08
Unless he was a Cubs fan to be
2:10
the opposite of his brother. because his brother is a Sox fan. Could be.
2:16
I mean, if all the trouble's in the world, and that's all we have to worry about
2:21
is what sports team he supports.
2:24
But, hey, it's Chicago.
2:30
My reaction, I was totally stunned because I'd never heard the name before,
2:37
had no idea who this man was.
2:41
And really was kind of, not kind of, very worried about who they were,
2:48
you know, who were they going to say it was.
2:52
I, you know, had my picks of who I thought.
2:56
But, you know, we're human. We're not the Holy Spirit. We didn't do this.
3:02
We could not have anticipated this. I don't believe.
3:06
And they come out and say, we have a pope. Then they say his name. And I thought, huh?
3:13
And it didn't take long, only four ballots. Right. So they must have had him
3:17
in mind, but I really hadn't really actually heard of him hardly before.
3:21
I hadn't either. It was just astonishing.
3:24
And then, you know, find out they say American.
3:28
And I thought, well, could it be Dolan?
3:33
Could it be, you know, somebody from the West Coast? I,
3:38
And here they come out and then they say his name. I don't know who that is. Yeah, Robert Previst.
3:44
Yeah. I'm not sure, you know. But he was a missionary.
3:49
He also taught at the Augustinian University, Villanova, in Pennsylvania. He taught.
3:58
His undergraduate work, he studied mathematics. That was his.
4:04
So maybe he's got a logical thinking.
4:07
Yeah, he's supposed to be an intellectual and very smart and so forth.
4:11
I hope that he is more precise in his pronunciations and his comments about
4:17
things than Francis was.
4:19
Francis often said things that were iffy and left things sort of hanging,
4:23
and you really didn't know for sure exactly what he intended or meant by him.
4:27
And I don't think that's good for the church, really. And I hope this man seems
4:32
to be maybe more careful.
4:35
And I think it's maybe they also wanted to bring someone in who would be a unifier.
4:43
Right. We really do need that. We need to be united.
4:48
We need some peace. Don't need all this push and pull in the church,
4:54
and nobody knows which way they're going. This has to settle down.
4:59
When he came out and he had on the papal vestments, boy, did that relieve my,
5:07
I was just relieved. Yeah. Oh.
5:10
I was too. I was glad to see him. Real pulp. Have him see the full vestments.
5:15
He even had on the red shoes as well. I didn't, you know, see that,
5:18
but I mean, they said that later. Yeah.
5:21
And you know what they symbolize? Right.
5:24
The red shoes. The martyrs. That you're walking on the blood of the martyrs.
5:31
And then his first mass, he said in Latin, I'm not sure if it was nervous.
5:40
I can't say that today. Novus.
5:44
It was in Latin. So that eased my mind, too.
5:50
And his first greeting, I thought, was very good.
5:55
So maybe we'll have a pope that will promote understanding, peace, and unity.
6:04
He speaks about five or six different languages. Oh, yeah. I think so. Even in the original...
6:10
Cancun or not Cancun. The original, is it Inca language? He knows.
6:20
I'm not sure. Yeah, I mean, he's quite versed in things. It's amazing.
6:26
But to have, you know, I just think this is going to put things at ease in the United States.
6:34
Not to have somebody pointing a figure at you and criticizing everything you do over here.
6:39
I think that's going to help because if he starts doing that we'll be shocked.
6:46
Well we'll see how his papacy works out and we'll certainly pray for him.
6:49
I'm certainly going to support him and pray for him.
6:52
He's got we got to give him a chance so we'll see what we shall see okay are you finished?
7:03
Yeah. Alright well I just want to be sure oh okay Okay, we're going to talk
7:10
then today a little bit about AI.
7:13
And I'm going to read from an article in First Things magazine, April 2025.
7:21
And it's by Thomas Fowler. And Thomas Fowler is president of the Xavier Zveri
7:28
Foundation, a technology consultant to the U.S.
7:31
Government, an adjunct professor of engineering at George Mason University.
7:35
He is author of four books and 150 articles on philosophy, theology,
7:41
engineering, mathematics, astronomy, and physics—,
7:44
In today's environment, he is especially interested in correcting widely promoted,
7:49
but in correct perceptions of science and its capabilities.
7:53
And there has been an awful lot of hype concerning AI and a lot of fear mongering as far as that goes.
8:02
I was just watching yesterday on TV. There was a man, Tristan Harris,
8:07
from the Center for Humane Technology.
8:10
And he sat there on the TV set looking like, with his eyes,
8:14
looking like a deer caught in headlights and a long face and predicting that in a couple of years,
8:21
AI will surpass human understanding and we could possibly have something like the Matrix, you know,
8:28
and the AI could take over and all of these things or AI can become human and
8:34
have a soul and do we have the right then to turn it off and all of the stuff that you saw in 2001,
8:39
a space odyssey with the computer HAL that was projected 50, 60 years ago.
8:45
And there has been from the very beginning about computers from 70 years ago
8:50
or even longer ago than that, all kinds of predictions and so forth that haven't come true.
8:55
And the fact is, is a lot of this is hype. And Thomas Fowler addresses that.
9:01
One article is in the magazine Touchstone, And that's one that incorporates
9:07
and concerns Christianity.
9:09
And I'm not going to have time to read much from that one today.
9:12
He has the other one in First Things, and I'm going to read from that one.
9:16
It's a little bit more concise, and I'll read some other things.
9:21
But the point is, is that there has been a lot of speculation about AI and so forth.
9:28
So let's take a look at it and see exactly what AI is, how it works,
9:33
and whether or not it has met all of the tests and so forth that people have
9:40
said that it is going to meet.
9:42
And there are a lot of these things about AI that are testable, and we'll see.
9:47
So anyway, it says the title of the article is AI Doesn't Know What It's Doing by Thomas Fowler.
9:56
He goes on, Artificial intelligence is a Numbrella term covering many beliefs
10:01
about the powers possessed by computers,
10:03
both now and in the future, because computers today perform many tasks formerly
10:10
reserved to humans many observers predict that they will soon replicate human
10:15
intelligence and gain greater capabilities thereafter,
10:19
finally rendering mankind obsolete.
10:22
Neither what AI actually is, nor the paradigm of knowing it employs,
10:27
is ever discussed by the AI believers.
10:30
Even though both questions are essential to understanding what AI can do and
10:36
what its limitations are.
10:38
As we shall see, the real dangers of AI are, ironically, byproducts of the hype
10:45
swirling around it, which attributes to it capabilities and reliability that it will never have.
10:52
In common parlance, artificial intelligence denotes computers that imitate people,
10:58
or computers that are just like human brains, only smarter.
11:02
There is no broadly accepted definition of AI, so let's formulate one.
11:07
AI is the category of systems that employ computers, feedback,
11:14
rule-based logical inferences, deterministic or statistically,
11:19
complex data structures, and large databases to extract information and patterns
11:25
from data and apply them to the control of equipment,
11:29
assistance with decision-making, or the generation of responses to user queries
11:35
involving texts and images.
11:37
I don't know if you can remember that.
11:40
The kinds of technology that typically fall under the rubric of AI include robots
11:46
and robotic systems, neural networks and pattern recognition.
11:51
Generative AI, including chat GPT, and similar large language models,
11:58
symbolic manipulation programs such as Mathematica, autonomous cars and other
12:04
autonomous systems, and complex large-scale control programs.
12:10
There are four fundamental questions about AI. One, what is theoretically possible for AI?
12:18
Number two, what is practically possible given current technology?
12:24
Number three, what is economically efficient for AI in terms of costs and benefits?
12:30
And finally, what ethical boundaries are appropriate for AI?
12:35
This article is concerned primarily with the first question.
12:40
Since it bounds the other three. First, we look at reasons for the hype,
12:45
and there is a lot of hype, believe me, surrounding AI.
12:49
Extravagant claims about computers have a long history, dating to the introduction
12:54
of the first commercial model, the UNIVAC-1, in 1951.
12:59
During the 1950s, computers were called electronic brains.
13:04
Computer pioneer Alan Turing, 1912 to 1954,
13:09
informed us 70 years ago, quote, it seems probable that once the machine thinking
13:16
method had started, it would not take long to outstrip our feeble powers.
13:21
And notice that phrase, our feeble powers.
13:23
Now, that's an asseveration, an assertion that needs to be looked at pretty hard.
13:30
If we take a look at all of the things that human beings have been able to do
13:34
on this planet and on other planets as well, more lately, our feeble powers
13:40
are not quite as feeble as maybe some scientists think they are.
13:44
Machines would be able to converse with each other, to sharpen their wits.
13:49
At some stage, therefore, we should have to expect the machines to take control.
13:53
This has been a theme for a long, long time concerning computer thinking.
13:59
In the film 2001, A Space Odyssey, 1968, the intelligent computer,
14:05
HAL, seizes control from the human astronauts.
14:10
Similar takeovers have been projected in warfare and in white-collar professions,
14:15
such as legal advice, financial consulting, and teaching.
14:19
Beyond that, we are told that computers will become conscious,
14:23
will develop full human capabilities, and who knows, may have souls.
14:27
And that's what this guy the other day on TV was talking about with the long
14:31
face and the deer-in-the-headlight eyes.
14:34
And a lot of that is simply over-exaggerated.
14:38
Actually, think about it. So you create a computer that can do all the things that you can do.
14:47
It can think, and they even think maybe they can create a soul.
14:52
What does that tell you? Man is foolish.
14:57
God creates.
14:59
Well, there are a lot of techies and scientists and so forth that reject creation
15:04
entirely in the first place. Well, yeah, they reject God.
15:06
Yeah, they don't like the idea of human beings being created.
15:09
At any rate, this is the viewpoint known as artificial general intelligence.
15:15
Machines will have intelligence similar in kind to human intelligence, but superior to it.
15:20
Ray Kurzweil has promoted the idea of a singularity, that is,
15:25
a merger between human intelligence and machine intelligence that is going to
15:30
create something bigger than itself.
15:32
The hype continues to escalate, particularly in the direction of threats posed by AI.
15:38
And there have been a lot of stories along this line all the way back from the
15:41
1950s up until now, and not just in 2001, a space odyssey, but in a lot of science
15:48
fiction stories and so forth about the machines and the computers taking over.
15:52
So far, none of those have seriously manifested themselves, but they still go on.
15:59
And these threats supposedly require immediate action to save humanity from some huge catastrophe.
16:07
So we're going to stop here and take a break. So stay with us. We'll be right back.
16:12
You've been listening to Being Catholic with Bob Johnston on Catholic Spirit
16:16
Radio. Tired of your phone bill supporting who knows what?
16:21
ICatholic Mobile is a 5G provider that supports what matters to you.
16:27
With iCatholic Mobile, every call, text, and even that endless group chat helps
16:35
support the mission of the Catholic Spirit Radio Network.
16:38
It's a network that connects you to what's good for both your soul and your wallet.
16:45
More information is available at iCatholicMobile.com.
16:50
ICatholic Mobile underwrites for the Catholic Spirit Radio.
16:55
Time is quickly running out for you to make your reservations for the August
16:59
pilgrimage to Alabama to see the EWTN studio and Mother Angelica's Shrine of
17:04
the Most Blessed Sacrament.
17:05
We hope you'll join us on this wonderful bus trip.
17:09
Registration closes on May 31st. Hurry! The details are found online at catholicspiritradio.com
17:15
or ask for Kathy at 309-828-6554.
17:24
Are you or someone you love seeking freedom from addiction, compulsions,
17:27
or unhealthy attachments?
17:29
Join Catholic in Recovery for a powerful day of healing at the Redemptorist
17:33
Retreat Center in Oconomowoc, Wisconsin, just outside Milwaukee on Saturday, June 7th.
17:39
Experience Christ's mercy through
17:41
the sacraments, prayer, and fellowship rooted in 12-step principles.
17:44
Register now at catholicinrecovery.com. That's catholicinrecovery.com.
17:50
You are not alone. Prayer for the Canonization of Archbishop Fulton J.
17:55
Sheen Written by Monsignor Richard Soseman Heavenly Father, Source of all holiness,
18:00
You raise up within the Church in every age men and women who serve with heroic love and dedication.
18:06
You have blessed Your Church through the life and ministry of Your faithful
18:10
servant, Archbishop Fulton J.
18:12
Sheen. He has written and spoken well of Your Divine Son, Jesus Christ,
18:16
and was a true instrument of the Holy Spirit in touching the hearts of countless people.
18:22
If it be according to your will, for the honor and glory of the Most Holy Trinity
18:26
and the salvation of souls, we ask you to move the church to proclaim him a saint.
18:32
We ask this prayer through Jesus Christ our Lord. Amen.
18:40
Hi, this is Bob Johnston. You're listening to Being Catholic right here on Catholic Spirit Radio.
18:46
We're back from our break, and we're talking about AI,
18:49
and in particular, we're talking about all of the doomsday threats that have
18:53
been made about computers taking over our society and taking over from human
18:59
beings and becoming gods and so forth,
19:01
and something like the Matrix and enslaving human beings that has been going on for the last 50, 60,
19:08
70 years, and is still going on today.
19:11
And all these threats require immediate action, according to some of these people.
19:16
We have to save humanity from catastrophe.
19:19
Some argue that artificial intelligence, could someday destroy America.
19:23
Others aver that catastrophe is right around the corner.
19:27
Eliezer Yarkowsky, a researcher at the Machine Intelligence Research Institute,
19:32
warns that the death of humanity is the obvious thing that would happen.
19:38
Jeffrey Hinton, the godfather of AI, recently assessed the likelihood of human
19:43
extinction due to AI as 10 to 20 percent.
19:46
And as I said yesterday, this one guy that was on TV talking about AI and predicting
19:54
similar doomsday scenarios was scaring the bejesus, as he said,
20:00
out of John Roberts, the newscaster who was interviewing him.
20:04
And the problem is that a lot of these people that report on this really are
20:09
very unaware and don't know much about AI and don't know much about computers and how computers work.
20:17
And they take a lot of this stuff literally, and they pass on that fear to people
20:22
watching who are in the same position.
20:24
And that's why we're addressing it, because as Catholics and as Christians and
20:29
Christianity, there's also the charge that AI is going to become a new God and
20:35
take over our society in a God-like manner and so forth.
20:38
And a lot of this, like I say, is hype, and we'll see why.
20:43
AI is feared for another reason, namely that it may be a disruptive technology,
20:49
one that causes major changes to areas of business, industry,
20:52
and commerce, thereby threatening the livelihood and normal activities of most of the population.
20:58
And that's a little bit more realistic, a more realistic threat.
21:02
We have had lots of disruptive technologies in our society that we've had to
21:07
deal with, and AI certainly could be another one.
21:10
The automobile and the personal computer are prime examples of disruptive technologies.
21:15
I mean, they have already replaced a lot of human beings in various works.
21:20
But for a technology to be disruptive, it needs to actually work.
21:24
So that's the thing with AI is, first of all, is it actually going to work as people say it is?
21:30
In this case, they interact with the world as humans do, only better.
21:34
It is unclear that AI will ever replicate human cognition.
21:40
Indeed, it is reasonable to ask whether in 70 years we have moved any closer to Turing's vision.
21:47
If computer power has vastly increased and computer size has shrunk dramatically,
21:53
what has it all accomplished?
21:55
The assumption is that such progress will lead eventually to qualitative changes in machine behavior.
22:02
This is an empirically testable proposition. In other words,
22:06
we can take a look and see, as machines have improved and have gotten more powerful,
22:11
have they come closer to human-taught cognition.
22:14
Comparison of a mainframe computer from Turing's day with a modern smartphone
22:20
shows improvements of 6 to 13 orders of magnitude.
22:24
In other words, a lot of computer phones are 13 times as powerful as the biggest
22:30
mainframe computers were back in the 1950s.
22:33
But do they show any evidence of sentience, that is, of self-awareness or the
22:40
ability to think like a human being? And the answer is no.
22:44
Plainly, it is taking longer to outstrip our so-called feeble powers than Turing envisioned.
22:50
No one regards a smartphone as anything more than a handy multipurpose tool.
22:55
We certainly don't expect it to take over our household or to start ruling us tomorrow.
23:00
Similar remarks can be made about quantum computers.
23:04
And these are computers that are extremely fast. And we'll look into that.
23:10
Will super fast computers do any better at this? We'll see.
23:15
The implication that the scaling
23:19
of computer power will not yield the outcomes postulated by Turing.
23:23
I mean, so far the evidence has shown that these increases in power have really
23:28
not yielded any more sentience or any step toward it on the part of computers than we had before.
23:36
ChatGP and other generative AI programs are popular but have established an
23:41
unenviable track record.
23:43
Let us consider some of their gaffes. Climate scientist Tony Heller asked ChatGP
23:49
a simple question about CO2 levels, corals, and shellfish.
23:54
The answer at return, beginning, quote, if atmospheric carbon dioxide levels
24:00
were to increase by a factor of 10,
24:02
it would have significant and potentially devastating impacts on corals and
24:06
shellfish, was completely wrong, a mindless echo of the climate alarmism that
24:12
is everywhere on the internet.
24:15
ChatGDP has been known to make up articles and bylines, a proclivity that has
24:20
struck the Guardian since phony articles have been distributed to it.
24:24
Quote, Huge amounts have been written about generative AI's tendency to manufacture
24:30
facts and events, But this specific wrinkle, the invention of sources,
24:35
is particularly troubling for trusted news organizations and journalists whose
24:40
inclusion adds legitimacy and weight to a persuasively written fantasy.
24:45
In other words, AI often generates phony articles.
24:48
Sometimes it's because the information that's being fed from the Internet,
24:52
especially, or other information is inaccurate. and it simply absorbs that inaccurate
24:58
information and spouts it back out.
25:01
But it also can generate false information.
25:05
And the reason it can do so is a computer crunches numbers.
25:10
Those numbers can be turned into words, and they are turned into words according
25:14
to certain grammatical algorithms that are fed into the computer.
25:18
The computer then can do millions upon millions, even billions of different
25:25
combinations of numbers and the words that follow those numbers.
25:29
So, eventually, it can write almost anything, and it does, because it doesn't know what it's doing.
25:36
A human being does research, and he knows what his purpose is.
25:39
He knows that he's doing it.
25:41
He searches for truth, and he has the ability with his cognitive mind to reach
25:46
even abstract truths and so forth, and he filters out a lot of things and decides
25:50
what is valid and what is not valid, what is useful and what is not useful, and so forth,
25:55
the chat box, the computers, and so forth, don't do this.
25:59
They do whatever they're programmed to do by the algorithms that are fed into them.
26:04
Recent research has shown that the chat box are getting worse at basic math,
26:09
as shown by their inability to answer reliably such questions as whether a given
26:14
number is prime. In other words, for.
26:18
A human being knows that three, five, and seven are prime numbers.
26:21
These are numbers that can be divided evenly only by one and by themselves.
26:25
And the human beings can recognize these numbers.
26:29
Computers can't do this. One of the reasons is computers have a very,
26:33
very hard time with universals, what we call abstractions.
26:37
They can't understand abstractions. In most literature, in most things that
26:41
they're fed to read are abstract in nature. I mean, when we read something,
26:46
a lot of times it means exactly what it says, but in a lot of cases, it doesn't do that.
26:51
It means what it says, and then it means something even further and deeper than that.
26:56
And sometimes it means the opposite of what it says if we're talking about satire
27:00
or things along those lines.
27:01
The computer has a hard time trying to recognize these things.
27:06
In another case, a lawyer used a chat box to research and write a legal brief.
27:11
Unfortunately for the lawyer, the brief contained numerous bogus legal decisions.
27:16
The computer just made them up.
27:17
And why can it do this? Well, it's programmed to.
27:20
And made up quotes leading to potential sanctions for the lawyer.
27:24
So he got into some serious trouble because he relied on the computer to give
27:28
him accurate information, and of course it didn't.
27:31
Obviously, if citations are untrustworthy and entire articles can be made up,
27:36
academic research, journalism, and everything in our society that depends upon
27:41
reliable knowledge will be undermined.
27:44
The New York Times has explored this problem, which strikes at the heart of
27:48
any notion of intelligence.
27:50
The Times asks ChatGPT a question, quote, when did the New York Times first
27:56
report on artificial intelligence?
27:58
The chat box answer invoked an article that the chat box has simply made up.
28:03
The inaccuracies that emerge from chat boxes and other such programs are called
28:09
hallucinations by those in the technology industry.
28:12
Serious research cannot be grounded on hallucinations.
28:16
Large language models like ChatGPT are based on analysis of enormous amounts
28:22
of data from various sources, usually the Internet.
28:26
The goal is to find patterns in the data, then construct text or images that
28:32
conform to these patterns.
28:34
Since the Internet contains much erroneous data, incorrect inferences,
28:39
and unbridled speculation, such an approach is highly problematic.
28:43
The Times observes, Because the Internet is filled with untruthful information,
28:49
the technology learns to repeat the same untruths.
28:52
And sometimes the chat box makes things up.
28:55
They produce new text combining billions of patterns in unexpected ways.
29:01
They don't know what they're doing. They're just doing what their algorithm tells them to do.
29:05
This means even if they learn solely from the text that is accurate,
29:10
they may still generate something that is not.
29:13
And if you ask the same question twice, they can generate different texts and different answers.
29:20
Even Microsoft has conceded that the chatbots are not bound to give truthful information.
29:26
According to an internal document quoted by the Times, AI is built to be persuasive.
29:31
It's not built to be truthful, with the consequences that outputs can look very
29:36
realistic but include statements that simply aren't true.
29:41
Chatbots are often misused by students who ask them for term papers or similar assignments.
29:48
An experienced teacher will easily perceive that the work is not the students
29:52
on the basis of style, sloppy reasoning, and bogus references.
29:58
But the imposter places an additional burden on the teacher.
30:02
Chatbots are the latest version of something that has been around for a long,
30:06
long time, and that is sophistry.
30:09
As the Eliatic stranger tells us in Plato's Sopis,
30:13
Now shouldn't we presume that there is some other skill involving words by which
30:19
one could beguile the young through their ears with words while they are still
30:23
at a far removed from matters of truth,
30:26
showing them verbal images of everything so as to make the statements seem true
30:31
and the speakers seem the wisest of all men on every issue?
30:35
The chatbots show once again that it is easy to pretend to knowledge,
30:40
but much, much harder to engage reality and truly know something about it.
30:45
A cautionary tale from the history of science is in order.
30:49
In the early days of telescopes, the quality of optothel glass was poor,
30:54
lens grinding methods were crude, and little was understood about what we now call physical optics.
31:01
As advancements came in all these areas, telescope performance naturally improved.
31:06
At the time, no limit to how good telescopes could be in terms of resolution
31:11
and color correction was foreseen.
31:14
Early telescopes makers did not understand the phenomenon of diffraction,
31:19
which limits the performance of any optical system, no matter how perfect the
31:24
lenses, mirrors, and other components.
31:27
So giddy as astronomers were about their better glass, improved lens grinding,
31:31
and innovative multiple lens objectives, they faced an unknown barrier that
31:36
ultimately would thwart their plants.
31:38
In other words, they thought that they would be able to get telescopes and so
31:41
forth that would be so good, you know, they could look at the moon,
31:45
so to speak, and see a dime or something sitting on its surface.
31:48
And they were just thinking that we'll be able to look at everything in the
31:53
universe and see it up close.
31:54
But that, of course, didn't happen because they didn't understand that there
31:58
were barriers to better and better telescopes, no matter how perfect they made them.
32:03
Long-range extrapolation of technology is likely to be met with disappointment.
32:08
Did you know that there were people back in the late 1800s in cities like Chicago
32:16
that predicted that Chicago would be covered with 40 feet of horseshit because of all the horses,
32:22
the city growing bigger and bigger and the horses moving in and they wouldn't
32:26
be able to remove it and they'd be covered in horse crap?
32:29
Of course, the automobile was right around the corner and that didn't happen.
32:34
And if these cities today are covered with horse manure, it's mostly because
32:40
of all of the statements of the politicians that occupy those cities today.
32:45
But modern AI is based on ideas of human knowing that stem from the British
32:50
empiricist tradition, in particular, the philosophy of David Holm.
32:55
Holman envisioned the body as a composite of discrete physical systems,
32:59
with the senses sending their reports to the mind, which then worked on these reports.
33:04
These reports he termed impressions, which gave rise to ideas.
33:08
In other words, he looked at the human body as sort of some kind of a telegraph
33:11
system, similar to a computer, with the outer areas of the human body sending
33:16
all kinds of reports to the brain and the mind part of the brain.
33:21
And then these reports were put together into ideas and so forth.
33:24
Sort of a mechanical process.
33:26
And I doubt if that's at all anywhere near even how human beings actually cognate and how they think.
33:33
And he says, I venture to affirm that the rule here holds without any exception
33:37
and that every simple idea has a simple impression which resembles it and every
33:43
simple impression a quorum exponent idea.
33:46
Holm presents a theory of knowing in which senses deliver impressions,
33:50
which we process as ideas.
33:52
Once we have ideas, we can reason with them, either by means of logical inference
33:57
or directly as matters of fact, empirically grounded facts, including scientific laws.
34:03
As for general ideas, they are nothing more than particular representations
34:08
connected to a certain general term.
34:12
This theory quickly leads to nominalism, the belief that abstract entities do
34:17
not exist, and that any talk of entity such as mankind refers only to collections of individuals.
34:24
In other words, a lot of materialists who try to explain everything in terms
34:29
of material world can't deal with universals, universals such as redness or
34:34
holiness or things along that line.
34:37
And universals are not material things, and they don't exist in the real world.
34:43
Triangularity, for instance, is a universal. We know that a perfect triangle
34:48
contains 180 degrees and three closed, absolutely straight lines.
34:52
But no perfect triangle exists in the real world.
34:56
In the real world, we have triangles that we draw on paper, triangles that we
35:00
make out of various materials, cardboard or wood or something else, or stone and so forth.
35:05
And none of them, no matter how carefully we measure and how good our tools
35:09
are, if we blew those up to a large enough size, we would see that there are
35:13
all kinds of errors in them.
35:15
So they don't really exist in the real world.
35:17
Yet our minds can conceive of a perfect triangle and conceive of something called triangularity.
35:24
And that triangularity exists somewhere.
35:26
And if it doesn't exist in the real world, Plato insisted that it existed in
35:32
what he called the third realm. So we're going to have to stop here and take a break.
35:37
So stay with us. We'll be right back. You've been listening to Being Catholic
35:40
with Bob Johnston on Catholic Spirit Radio.
35:44
Time is quickly running out for you to make your reservations for the August
35:47
pilgrimage to Alabama to see the EWTN studio and Mother Angelica's shrine of
35:52
the Most Blessed Sacrament.
35:54
We hope you'll join us on this wonderful bus trip.
35:57
Registration closes on May 31st. hurry. The details are found online at catholicspiritradio.com
36:03
or ask for Kathy at 309-828-6554.
36:12
Are you or someone you love seeking freedom from addiction, compulsions,
36:16
or unhealthy attachments?
36:17
Join Catholic in Recovery for a powerful day of healing at the Redemptorist
36:21
Retreat Center in Oconomowoc, Wisconsin, just outside Milwaukee on Saturday, June 7th.
36:27
Experience Christ's mercy through
36:29
the sacraments, prayer, and fellowship rooted in 12-step principles.
36:33
Register now at catholicinrecovery.com. That's catholicinrecovery.com.
36:38
You are not alone. Prayer for the Canonization of Archbishop Fulton J.
36:43
Sheen Written by Monsignor Richard Sosman Heavenly Father, Source of all holiness,
36:49
You raise up within the Church in every age Men and women who serve with heroic love and dedication.
36:55
You have blessed Your Church through the life and ministry Of Your faithful
36:58
servant, Archbishop Fulton J.
37:00
Sheen. He has written and spoken well of Your Divine Son, Jesus Christ,
37:05
And was a true instrument of the Holy Spirit In touching the hearts of countless people.
37:09
If it be according to your will, for the honor and glory of the Most Holy Trinity
37:14
and the salvation of souls, we ask you to move the Church to proclaim Him a saint.
37:20
We ask this prayer through Jesus Christ our Lord. Amen. iCatholic Mobile is
37:25
a proud sponsor of Catholic Spirit Radio.
37:27
What if your phone bill did more than just connect calls? What if it connected souls?
37:32
With iCatholic Mobile, every text, every call, and every gig of data supports
37:37
Catholic Spirit Radio nationwide.
37:40
Y5G, unlimited talk and text, reliable data, and free roaming in Canada and
37:45
Mexico, all in a network that puts faith first.
37:49
More info at icatholicmobile.com.
37:57
Hi, this is Bob Johnston. You're listening to Being Catholic right here on Catholic Spirit Radio.
38:02
We're back from our break, and we're talking about AI, and we're talking about
38:07
some of the exaggerations about AI and the threats that,
38:11
supposedly makes a society. And we're talking a little bit about how actually
38:16
the human mind works in comparison to AI, how AI works and how the difference is between the two.
38:23
And we're talking here about as soon as you try to explain the mind in the same
38:28
way that you explain a computer in a materialistic way, you run into all kinds of problems.
38:34
And a lot of these techies and scientists simply dismiss these problems and
38:37
pretend they don't exist.
38:39
And one of them is this, As soon as you start talking about the mind or even
38:43
talking about language, you start running into what we call in philosophy what
38:48
is called universals, that is, abstractions that we use in order even to speak.
38:54
In fact, the sciences uses these abstractions in order to do science,
38:57
and yet a lot of scientists will say these abstractions don't really exist or
39:02
that they exist, but they only exist in the human mind.
39:05
The problem is, is that simply isn't true.
39:08
The fact is, is the human mind knows what triangularity is, even though in the
39:13
real world, no perfect triangularity exists.
39:16
And the real triangles in the real world are far from the perfect triangularity
39:20
that exists in our minds.
39:22
And the fact is, is it's pretty easy to make the logical thought that if all
39:28
human beings went out of existence tomorrow, that doesn't mean that triangularity goes out of existence.
39:34
180 degrees would still equal three closed sides.
39:37
And the Pythagorean theorem, the idea that the hypotenuse of a right triangle
39:43
is equal to the sum of the squares of the two sides, would still be in existence.
39:47
And those things would exist in any world that was in existence.
39:52
In fact, our minds can know this.
39:53
We know, for example, that unicorns may or may not exist in a given world.
39:58
Maybe on some planet somewhere a unicorn exists. We don't know.
40:02
It could exist, and it might not exist.
40:04
But we do know that certain things have to exist.
40:07
We would know that triangularity would exist no matter where in the universe it was.
40:12
We would know, for example, that 2 and 2 equals 4, and it doesn't equal 23.
40:16
And that exists anywhere. So, there are certain things, abstractions,
40:20
that exist in a human mind, but nevertheless would exist even if the human mind was gone.
40:25
And we know that those things don't really exist in the real world,
40:28
and yet they're real, so where is it that they do exist?
40:31
Plato said that they exist in the third realm, wherever that is,
40:34
and of course, most Christian thinkers conclude that those things exist in the
40:39
mind of God, and the human mind, the imago Dei, is made in the image of God.
40:46
Holm was never to explain how we arrive at forms of knowledge such as science,
40:50
mathematics, and history.
40:52
What impression gave rise to Einstein's field equations for general relativity?
40:57
Because every idea must be associated with the precedent impression resembling it,
41:02
Holm could not explain how we can do something as simple as recognize a thing
41:06
that is in a different position than when we first saw it, a problem that bedevils
41:11
AI systems used in autonomous cars.
41:14
In other words, computers can't do this. A human child can take a look at something
41:18
once and take a look at a cat or a car and then see the same thing a day or
41:22
two later in a different position, upside down or sideways, and still recognize it for what it is.
41:27
Computers can't. They have to be taught every single position there is.
41:31
Nor was he able to explain how it is possible to have knowledge of almost anything
41:36
without recognizing abstract entities is real.
41:39
For example, the statement, Beethoven's fifth is a great symphony,
41:45
uses abstract entities as both subject and predicate.
41:48
Had there never been any performance of the notes Beethoven wrote,
41:52
the statement would still be meaningful and true.
41:54
And the term great symphony refers not to a collection of performances of music,
42:00
but to a real characteristic of a certain type of musical composition.
42:04
To demonstrate AI's indebtedness to Holmes' theory of knowing,
42:08
let's consider two implementations along with some of the problems that pertain
42:12
to Holmes' theory and hence also to AI.
42:16
In robotic systems, which include robots and self-driving cars,
42:20
sensors send reports to a central processor, which employs algorithms to do
42:25
calculations on the data and then instructs mechanical parts to carry out certain operations.
42:32
Robotic systems accept Holmes' notion of the separability of sensing and knowing.
42:37
They emulate it in accordance with the standard engineering practice of isolating system functions.
42:44
The ideas are software structures that arise from impressions given by sensors.
42:50
The system reasons by means of software manipulations applied to these ideas.
42:56
The system is nominalist because it has no concept of abstract entities,
43:00
only of the concrete objects in front of it.
43:03
In other words, it can't recognize universals, can't recognize,
43:07
for example, there is something that exists called redness, or something that
43:11
exists called holiness, or something that exists called trees or treeness.
43:16
Though generative AI is structured differently than AI applications such as
43:20
robotics, it shares with them one key assumption, namely nominalism.
43:26
Generative AI scans large collections of works that employ key words and phrases,
43:31
takes the results, and assembles them on the basis of frequency,
43:35
into a report following rules of grammar and knowledge of word order frequencies
43:40
without knowledge of the abstract entities and ideas involved.
43:44
In other words, it engages in a highly superficial form of reading.
43:49
The problem, and it is a problem that vitiates the entire approach,
43:53
is that most important texts can't be read this way.
43:57
Only in some cases, such as scientific and most historical writing,
44:01
is the literal meaning of a text its principal meaning or the only meaning.
44:05
For many works, especially works of literature and philosophy,
44:09
the message or theme requires a holistic understanding of the text.
44:13
It is not conveyed by any piece or excerpt that AI can scan.
44:18
Often, indeed, the meaning of a work may depend on the reader's imaginative
44:23
reception of it, as is the case with poetry.
44:26
And many texts have multiple levels of meaning, so that a literal reading may
44:31
be true, as far as it goes, while being less important than the symbolic reading.
44:36
Or the real meaning of a text may be the exact opposite of what it says.
44:41
Of its surface meaning, as in satirical writing.
44:44
In satire, you're saying one thing and meaning exactly the opposite. Thank you.
44:48
The purpose of much theological and poetic writing is to open a window onto
44:53
a numinous world, that is, a spiritual world, and texts in disciplines such
44:57
as philosophy may depend entirely on abstract ideas and entities.
45:02
The reader of any of these kinds of texts must be able to perceive the reality
45:07
behind the words, reading and understanding the entire text,
45:12
including very abstract ideas and what they entail or imply,
45:16
taking into consideration the writer's goal, his presuppositions and biases,
45:21
and then relating the work to others in order to ascertain its thoroughness,
45:26
accuracy, and contribution value.
45:28
Only thus can a thorough view of the subject emerge.
45:31
On this ground alone, it is plain that large language models will never replace
45:37
or replicate human knowing.
45:39
AI can parrot what real minds have thought and said on these topics,
45:43
and thus it can sound intelligent, what it cannot do is understand material.
45:49
Humans are distinguished by our ability to perceive reality.
45:53
Those of us who consistently misperceive reality are regarded as incapacitated, that is, mentally ill.
46:01
Our judicial system is based on the perception of reality, for it presumes real
46:06
crimes committed by real people
46:08
using real weapons and a perception or detection of all these realities.
46:13
Further, we humans believe that what we perceive directly may point to a reality
46:19
that is beyond perception.
46:21
Peter Kruft's famous argument is one example. There is the music of John Sebastian Bach.
46:27
Therefore, there must be a god. Computer could not understand that if it read it.
46:32
The ability of visual art to convey depth beyond the canvas is another.
46:36
In other words, you can look at a canvas in a painting and you can see depth in it.
46:40
You You can see distance and so forth, the computer can't.
46:43
The goal of human knowing is always to know something about reality,
46:47
regardless of whether that knowledge has operational value.
46:51
By contrast, neither an animal nor an AI seeks the reality of the real.
46:57
AI must employ symbols which have no meaning except that assigned to them by
47:02
someone outside the computer system.
47:05
The implication for the uniqueness of humans is straightforward.
47:10
Those who would assimilate humans to computers use an argument with this compound
47:14
premise, that humans are material only, and human functions can be reduced to algorithms.
47:20
The conclusion is, computers can duplicate human minds.
47:25
But if computers cannot duplicate human minds, and we can see that they can't,
47:30
then it follows that either or both humans are not material only,
47:35
or human functions cannot be reduced to algorithms.
47:38
And it's likely that humans are not material only, that our minds are not material.
47:43
Our brains may be, but our minds are not.
47:46
These are fairly momentous points, and they suggest one reason why understanding
47:51
what computers can and cannot do is important.
47:54
Human knowing operates on a principle that is radically different from A.I.'s homeon paradigm.
48:00
Humans know by means of an integrated system of sensing, motor skills, and the brain.
48:05
We have direct contact with reality, and we are able to know realities that
48:10
exist beyond the realities we immediately perceive.
48:14
In other words, we can understand things philosophically and so forth that are
48:19
beyond just the material reality that we do perceive.
48:23
This form of knowing is supremely creative.
48:26
It encompasses the way in which we understand situations we have never encountered
48:30
and generate new theories about reality.
48:34
Humans can think outside the box. AI cannot.
48:37
It has to stay within the premises and algorithms that it's given by whoever programs it.
48:43
AI can, of course, generate ideas understood in rather limited sense of data
48:49
structures or random chat box statements.
48:52
That is not how humans develop new theories or deal with unexpected situations.
48:58
Our perception of reality is unlike anything that can be achieved by any paradigm
49:03
based on separation of functions.
49:06
AI algorithms cannot creatively and analytically think through a question,
49:12
using information learned from reading and research,
49:15
bringing to bear a critical eye for discerning what is valuable,
49:19
and a perception of reality for synthesizing new ideas.
49:23
A human can do all this, and at the same time the human being knows who he is
49:26
and knows that he's doing it, the computer does it.
49:29
They can only ape human intelligence. The AI paradigm reacts to stimuli in the
49:35
form of sense-type data or website texts.
49:38
It cannot react, except very indirectly, to any underlying reality.
49:44
It does not know what it is doing.
49:46
Moreover, AI systems are backward-looking rather than forward-looking because
49:52
they are based on existing knowledge.
49:54
None has the ability to create new visions of reality, new theories.
49:58
Human beings can do this. Human beings use the material from their past,
50:03
things that they have learned from experience in observing the world,
50:08
and then they project those things forward into the future, and they can deal
50:13
with situations that they have never been in before. The computer can't.
50:17
None has the ability to create new visions of reality, new theories.
50:22
Of course, they can be used to make predictions or forecasts about the future.
50:26
Even simple regression analysis can do that. And they can help us to see things
50:31
that we otherwise could not see, such as simulations of the evolution of the universe.
50:37
But these simulations are based on current theories, for instance,
50:41
about the constitution of the universe and the laws governing it.
50:45
AI cannot advance human knowledge in any theoretical sense.
50:49
AI will be expected to do things that it will never be able to do.
50:53
The result will be fruitless expenditures of money and time.
50:56
Worse, AI-controlled systems may misbehave, leading to disaster.
51:01
In the recent past, computers took over manual-intensive tasks,
51:05
such as bookkeeping, account balancing, and stock transaction processing,
51:09
displacing legions of clerks.
51:12
No one today would transact with a bank that had dozens of people working adding
51:16
machines in the back room.
51:18
Newer computer systems will tackle more complicated but still well-defined tasks
51:23
which will likewise display some type of workers.
51:26
AI is suited to tasks that can be narrowly defined and implemented algorithmically.
51:33
Tasks that require spontaneous decision-making in uncertain environments.
51:37
In other words, most tasks, are very ill-suited to it.
51:42
What good, then, is AI? AI is an evolutionary development, a continuation of
51:48
the ongoing process of determining what is needed, then creating and improving
51:52
software and systems that meet these needs.
51:55
Today's word processors are a great advance upon the crude, character-based
51:59
processors of the first PCs.
52:02
Likewise, image editing and manipulation has come a long way,
52:06
and the latest AI-based versions continue that trajectory.
52:10
In the area of pattern recognition, important in medicine and other applications.
52:15
AI will result in improvements.
52:17
AI's real value will be in specialized applications where it can enhance human
52:22
capabilities and productivity.
52:24
AI will never become conscious, replace humanity, or take over.
52:29
It will displace some workers, though historically technology has created more
52:34
jobs than it destroys. It will therefore impose burdens on society to ensure
52:39
that those who are displaced are not abandoned.
52:42
Can AI be made smarter and overcome its currency deficiencies? Unlikely.
52:47
And I'm going to talk here about, in another article in Touchstone magazine,
52:52
a man talks about the fact that he worked on and was able to help build a supercomputer
52:58
that does one quintillion calculations per second.
53:04
Imagine that, one quintillion calculations per second.
53:07
In other words, if the whole human race on the planet Earth were doing those
53:11
calculations, It would take four years for all of the people working together
53:16
to do what this computer does in one second.
53:20
Did the computer become more aware of itself?
53:23
Did it take a step towards sentience? Did it take a step becoming like a human
53:29
being? And the answer is no, it didn't.
53:32
And the man himself points out that even the supercomputers are nothing but
53:36
rapid calculating machines.
53:37
And they can do a lot of things that they're programmed to do,
53:40
and that's all. They can't really think in terms of human thinking,
53:44
and they can't really be self-aware and know what they're doing and understand what they're doing.
53:49
And they still, as fast as those calculations are, they can be translated from
53:53
numbers into words, and they can write all kinds of things.
53:57
But those things are simply things that are written following certain programs
54:01
programmed into the computer. They're not anything creative.
54:05
In its AI usage, reasoning means committing fewer mistakes when dealing with
54:09
real-world problems by doing some fact-checking, which in turn amounts to taking
54:15
more time to weigh different scenarios.
54:17
This improvement does not alter the basic capabilities and limitations of LLMs,
54:24
large language machines, and computers in general.
54:28
At any rate, we're going to have to stop here. AI research are trying to square the circle.
54:33
If you begin with bad or inconsistent data, even a vast amount of statistical
54:39
knowledge cannot make it into something that is correct.
54:43
And they can't do it in a reliable manner.
54:45
So no, the answer is that computers will never do what human beings are able to do.
54:52
And the use that we have of computers will be based on factors of how expensive
54:57
are they, how much do they cost to run, how much electricity do they use, and so forth.
55:02
And they will be a tool just like they are a tool now, but they certainly will
55:06
not take over humanity and they will not take over the world.
55:10
So we will go ahead and say our prayer.
55:14
And this is the prayer that Leo XIII gave us.
55:18
And now we have a Pope Leo XIV who can continue on with that prayer.
55:24
St. Michael, the archangel, defend us in battle.
55:27
Be our protection against the wickedness and snares of the devil.
55:31
May God rebuke him and humbly pray.
55:34
And do thou, Prince of the Heavenly Host, by the power of God,
55:37
thrust into hell, Satan, and all evil spirits who wander to the world for the root of souls.
55:44
Amen. You've been listening to Being Catholic with Bob Johnston on Catholic
55:48
Spirit Radio. If you'd like to contact Bob, email bob at catholicspiritradio.com.
55:55
Again, that's bob at catholicspiritradio.com. Catholic Spirit Radio relies on
56:00
your support to bring programming like this and EWTN 24 hours a Please help
56:05
keep Catholic Spirit Radio on the air with your generous support.
56:08
Donate online at catholicspiritradio.com or send a donation to Catholic Spirit
56:13
Radio, 108 Boykins Place, Normal, Illinois, 61761.
56:18
That's Catholic Spirit Radio, 108 Boykins Place, Normal, Illinois, 61761.
56:23
Catholic Spirit Radio is a 501c3 and all donations are tax deductible.
56:28
Thank you for your support of Catholic Spirit Radio.